As AI evolves from Generative to Agentic, structural computing bottlenecks shift away from GPU and heavily toward CPU and system memory.
Predictor: Morgan Stanley
Prediction text
As AI evolves from Generative to Agentic, structural computing bottlenecks shift away from GPU and heavily toward CPU and system memory. | Server CPU TAM + agentic workload profiling
Key catalyst: Server CPU TAM + agentic workload profiling
Watch events: CPU utilization % in agentic workflows; server CPU TAM re-rating by sell-side.
Resolution evidence
Agentic workflows in production (Anthropic Claude agents, Cognition Devin, OpenAI agents) show measurable CPU-bound latency patterns.
Predictor: Morgan Stanley
Calibration plot (stated vs observed)
Evidence about this node from Morgan Stanley is multiplied by κ in /api/intake. Lower κ = less weight; floors at 0.10 (effectively silenced) and caps at 1.00 (full weight).
Reference class
This node isn't linked to a reference class. The Bayesian update applies without outside-view blending.
Probability over time
Milestone chain
- 2026-03-07overdueQ1 window check-in (25%)
- 2026-04-20hitMorgan Stanley publishes formal Rise of AI Agents report — GPU-to-CPU pivot thesisHow: Morgan Stanley publishes major research report formally identifying agentic AI shift from GPUs to CPUs and memory as structuralSource: Bitget News / Morgan Stanley — The Rise of AI Agents (April 2026)conf 99%
- 2026-05-12overdueQ2 window check-in (50%)
- 2026-04-22hitCPU-side orchestration is empirically measured at 50-90% of agentic workload latencyHow: Published benchmark or analyst report empirically measures CPU-side orchestration as 50-90% of agentic workload latencySource: ANI News — Agentic AI shifts value from GPUs to CPUs and memoryconf 90%Notes: Already cited by Morgan Stanley April 2026 — empirical CPU-bottleneck data live.
- 2026-07-17pendingQ3 window check-in (75%)
- 2026-04-01 → 2026-12-31pendingHyperscaler greenfield CPU racks deployed exclusively for agentic AIHow: At least one hyperscaler (Meta, Google, Microsoft, AWS) publicly discloses standalone CPU rack deployment dedicated to agentic AI orchestrationSource: BigGo Finance — Morgan Stanley: AI Agents Drive CPU Demandconf 85%Notes: NVIDIA standalone CPU at Meta already counts as initial signal.
- 2026-12-01 → 2027-12-31pendingMemory and CPU vendors outperform GPU pure-plays in 2026-27 returnsHow: Memory makers (Micron, SK Hynix, Samsung) + server CPU vendors (Intel, AMD, Arm) cumulative 12-mo total return exceeds NVIDIA over 2026-27 in any 12-mo windowSource: Morgan Stanley — Why Doubling Down on Memory Stocks Amid AI Boomconf 50%
- 2027-01-01 → 2030-12-31pendingDRAM demand from agentic workloads adds 15-45 EB by 2030How: Morgan Stanley / IDC track agentic-driven incremental DRAM demand reaching 15-45 exabytes by 2030Source: Morgan Stanley — Agentic AI shifts value from GPUs to CPUs and memoryconf 70%
- 2028-01-01 → 2030-12-31pending$32.5-60B incremental CPU TAM materialized by 2030 per Morgan StanleyHow: Server CPU TAM grows by ≥$30B incremental over 2026 baseline by 2030 attributable to agentic AI workloadsSource: NewKerala — Agentic AI Creates $60B CPU Market by 2030: Morgan Stanleyconf 65%
What if this resolves?
Click a button to clamp this prediction and run a Gibbs sample. Returns the predictions whose marginals shift most. ~30s per run; ideal for stress-testing "if X resolves, what else moves?"
Evidence chain
Raw metadata
{
"trf": 0.5881123843731557,
"kappa": 0.6333,
"base_rate": null,
"predictor": "Morgan Stanley",
"total_llr": -0.4054651081081644,
"grace_days": 7,
"bayesian_v2": true,
"prior_logit": -0.3657020103529956,
"bayes_factor": "1.3:1 against",
"blend_reason": "no reference_class linked",
"inside_prior": 0.409579974194961,
"kappa_source": "predictor_table",
"n_milestones": 1,
"blend_applied": false,
"contributions": [
{
"llr": -0.4054651081081644,
"kind": "quartile_checkpoint",
"kappa": 0.6333,
"label": "Q2 window check-in (50%)",
"weight": 0.05,
"strength": "weak",
"confidence": null,
"source_url": null,
"adjusted_llr": -0.2567810529649005,
"expected_date": "2026-05-12",
"measurement_criterion": null
}
],
"evidence_kind": "metadata_milestone_miss_sweep",
"inside_source": "history_v2",
"inside_weight": 0.5883213309387909,
"outside_weight": 0.4116786690612091,
"posterior_prob": 0.3492169282993375,
"posterior_logit": -0.6224830633178962,
"predictor_brier": 0.0442,
"inside_posterior": 0.3492169282993375,
"blended_posterior": 0.3492169282993375,
"reference_class_id": null,
"total_adjusted_llr": -0.2567810529649005,
"predictor_n_resolved": 2
}Raw metadata
{
"trf": 0.6650500679109432,
"kappa": 0.5833,
"base_rate": null,
"predictor": "Morgan Stanley",
"total_llr": -0.4054651081081644,
"grace_days": 7,
"bayesian_v2": true,
"prior_logit": 0.9444616088408513,
"bayes_factor": "1.3:1 against",
"blend_reason": "no reference_class linked",
"inside_prior": 0.72,
"kappa_source": "predictor_table",
"n_milestones": 1,
"blend_applied": false,
"contributions": [
{
"llr": -0.4054651081081644,
"kind": "quartile_checkpoint",
"kappa": 0.5833,
"label": "Q1 window check-in (25%)",
"weight": 0.05,
"strength": "weak",
"confidence": null,
"source_url": null,
"adjusted_llr": -0.2365077975594923,
"expected_date": "2026-03-07",
"measurement_criterion": null
}
],
"evidence_kind": "metadata_milestone_miss_sweep",
"inside_source": "prior_prob",
"inside_weight": 0.5344649524623397,
"outside_weight": 0.4655350475376603,
"posterior_prob": 0.6699488693573884,
"posterior_logit": 0.707953811281359,
"predictor_brier": 0.01,
"inside_posterior": 0.6699488693573884,
"blended_posterior": 0.6699488693573884,
"reference_class_id": null,
"total_adjusted_llr": -0.2365077975594923,
"predictor_n_resolved": 1
}Network propagation neighbors
No propagation data yet. Run inference/.venv/bin/python scripts/ops/run_loopy_belief_propagation.py on the droplet, or wait for the Sunday 02:00 UTC weekly cron.
Ticker exposure
Beneficiaries (22)
Prerequisites (1)
| Type | Pred | Title | Domain | Lag |
|---|---|---|---|---|
| correlate | S_COMPUTE_100GW_2030 | Compute: 100GW national-scale by Dec 2030 | compute_scale | — |
Dependents (0)
| Type | Pred | Title | Domain | Lag |
|---|---|---|---|---|
| No dependents | ||||
Validations (1)
| Observed at | Status | By | Notes |
|---|---|---|---|
| 2026-04-29 | partial | thesis_timeline_v1.0_import | Agentic workflows in production (Anthropic Claude agents, Cognition Devin, OpenAI agents) show measurable CPU-bound latency patterns. |
Linked documents (10)
Raw metadata
{
"nia": false,
"qty": "GPU→CPU shift",
"mode": "THESIS",
"role": "Cited-Firm",
"context": "Core thesis of MS 73-page 'Rise of the AI Agent – Global Implications' report; contrarian vs consensus GPU-centric view.",
"to_year": 2030,
"cited_by": "Rise of AI Agent report",
"conv_cues": "conclude; heavy thesis weight",
"direction": "HAPPEN",
"from_year": 2026,
"timeframe": "2026+",
"conv_level": "HIGH",
"milestones": [
{
"kind": "quartile_checkpoint",
"label": "Q1 window check-in (25%)",
"status": "overdue",
"weight": 0.05,
"ordinal": -6,
"source_id": null,
"expected_date": "2026-03-07",
"observed_date": null,
"miss_emitted_at": "2026-05-02T22:07:21.384228+00:00",
"miss_emitted_by": "metadata_milestone_sweep"
},
{
"kind": "llm_pre_event",
"label": "Morgan Stanley publishes formal Rise of AI Agents report — GPU-to-CPU pivot thesis",
"source": "Bitget News / Morgan Stanley — The Rise of AI Agents (April 2026)",
"status": "hit",
"weight": 0.4,
"ordinal": -5,
"source_id": null,
"confidence": 0.99,
"source_url": "https://www.bitget.com/amp/news/detail/12560605375963",
"expected_date": "2026-04-20",
"observed_date": "2026-04-20",
"research_origin": "deep_research",
"measurement_criterion": "Morgan Stanley publishes major research report formally identifying agentic AI shift from GPUs to CPUs and memory as structural"
},
{
"kind": "quartile_checkpoint",
"label": "Q2 window check-in (50%)",
"status": "overdue",
"weight": 0.05,
"ordinal": -4,
"source_id": null,
"expected_date": "2026-05-12",
"observed_date": null,
"miss_emitted_at": "2026-05-30T22:15:00.756418+00:00",
"miss_emitted_by": "metadata_milestone_sweep"
},
{
"kind": "llm_pre_event",
"label": "CPU-side orchestration is empirically measured at 50-90% of agentic workload latency",
"notes": "Already cited by Morgan Stanley April 2026 — empirical CPU-bottleneck data live.",
"source": "ANI News — Agentic AI shifts value from GPUs to CPUs and memory",
"status": "hit",
"weight": 0.4,
"ordinal": -3,
"source_id": null,
"confidence": 0.9,
"source_url": "https://aninews.in/news/business/morgan-stanley-agentic-ai-shifts-value-from-gpus-to-cpus-and-memory-creating-up-to-60bn-incremental-cpu-tam-by-203020260422131744/",
"expected_date": "2026-07-01",
"observed_date": "2026-04-22",
"research_origin": "deep_research",
"expected_date_range": {
"to": "2026-09-30",
"from": "2026-04-01"
},
"measurement_criterion": "Published benchmark or analyst report empirically measures CPU-side orchestration as 50-90% of agentic workload latency"
},
{
"kind": "quartile_checkpoint",
"label": "Q3 window check-in (75%)",
"status": "pending",
"weight": 0.05,
"ordinal": -2,
"source_id": null,
"expected_date": "2026-07-17",
"observed_date": null
},
{
"kind": "llm_pre_event",
"label": "Hyperscaler greenfield CPU racks deployed exclusively for agentic AI",
"notes": "NVIDIA standalone CPU at Meta already counts as initial signal.",
"source": "BigGo Finance — Morgan Stanley: AI Agents Drive CPU Demand",
"status": "pending",
"weight": 0.4,
"ordinal": -1,
"source_id": null,
"confidence": 0.85,
"source_url": "https://finance.biggo.com/news/-ffAq50BvthpMgHB4o2T",
"expected_date": "2026-08-16",
"research_origin": "deep_research",
"expected_date_range": {
"to": "2026-12-31",
"from": "2026-04-01"
},
"measurement_criterion": "At least one hyperscaler (Meta, Google, Microsoft, AWS) publicly discloses standalone CPU rack deployment dedicated to agentic AI orchestration"
},
{
"kind": "event",
... (truncated)